We propose to investigate the feasibility of an instrument to screen for obstructive sleep apnea syndrome (SAS). The instrument's operation will be based on an experimentally observed and published relationship between sleep apnea and speech disorders. This relationship is often visually detectable in the spectrogram of a person's speech. The spectrogram can be calculated in a few seconds with only the computing power of a standard personal computer. The envisioned instrument will appear physically as a common personal computer augmented by a microphone. Patients will pronounce a prescribed list of words into the microphone and a spectrographic analysis will be calculated and translated through a knowledge base, producing a quantitative estimate of the likelihood that the patient has the most common form of apnea (obstructive). If technically successful, the instrument would be highly attractive for use by hospitals and individual and group pulmonary and neurological practices, since it would provide an inexpensive, on-the-spot, non-invasive estimation of the existence of obstructive apnea, thus avoiding the need for an all-night sleep study in a significant number of potential SAS cases. Convenience, lower total cost, and reduced anxiety by reason of quick test results are the primary benefits foreseen for the screening instrument.